Liz Centoni, Cisco’s executive vice president and chief customer experience officer, sat down with CRN to discuss the company’s AI strategy with Splunk and how Cisco plans to use AI to power an inclusive future for all.
Liz Centoni, Cisco’s executive vice president and chief customer experience officer, has long been a champion of the tech giant’s most emerging, cutting-edge applications and technology, and artificial intelligence is no different. Centoni sat down with CRN to talk about Cisco’s AI strategy with Splunk, how partners can position Cisco AI at a time when the competition from networking and security players alike is at an all-time high, and how Cisco plans to use AI to power an inclusive future for all.
Cisco has referred to its AI strategy as a commitment, not a campaign. What does that mean?
When you think about deploying AI, which has huge opportunities whether you’re building tech for AI, you’re deploying it internally [or] you’re packaging it as part of your portfolio for customers to deliver high-value use cases, we always put privacy, security, ethics, reliability [first]. It’s things we’ve always done before—we had a responsible AI framework that was formalized in our company long before generative AI came out [and] long before what ChatGPT did in terms of democratizing access for everyone. So, as we think about AI, we’ve taken a purposeful stand that it must align with our purpose on powering that inclusive future for all and, as you can imagine, that’s dynamic. It’s not something that’s one and done. Just as the technology is so dynamic and [there are] lots of opportunities, [there are] lots of challenges along the way. It seems like a new surprise every single day.
The second thing is, while everybody talks about AI, it’s something that we've been working on for quite some time, especially when you think about the different branches of AI around machine learning [ML]. ML has been part of our portfolio for such a long time when you think about everything that we’ve done in networking around automation and assurance. This is volumes of data telemetry that we actually have in networking and security and in collaboration and observability. But when we talk about automation, assurance of the network, dynamic baselining—these are ML models we’ve run for quite some time, including things like computer vision, predictive analytics and recommendations [for] helping you decide as to whether you’re overprovisioning your applications. These are all things that were running in the background. With Gen AI and ChatGPT, an awesome set of tools, and now with multimodality, we continue to look at how do we leverage some of this capability but to deliver new use cases of value to our customers. And, how we leverage it internally as well. Internally, we look at it in terms of automated inspiration. We think about it as co-creation, not just this magical being that you have that will do everything for you, but there are some pretty magical things in terms of what you can co-create along with the tool. So, think about it as the human and the machine coming together, and that’s the way we absolutely look at it.
How will Splunk help Cisco become a disrupter in AI?
Our thesis for the acquisition, and we’ve been very public about this, is Splunk fits in very well with two core parts of our portfolio: security and observability. In security, we can take our customers to the next generation of what the future of a SOC [Security Operations Center] would look like. In the observability space, we come at this with AppDynamics [with] application performance monitoring, and [Splunk] has done this via log monitoring. So, as you look across all of this, and you think about the data that Cisco has, and then you bring in the vast amount of data that Splunk brings and the platform, it gives us the ability to provide richer insights than what we’ve been individually doing, and AI is going to be a key part of that. Think about summarizations as one example. You’re now providing summarizations based on, not just the data that Cisco has, but the data that Splunk potentially has as well. The richness of data that we can provide and the insight that allows us to prevent attacks and reduce the blast radius of attacks that are out there. And in observability, [we can] reduce the mean time to resolution to where we can get to predictive insight in both of those areas. I think there are very few peers out in the industry who can actually do that.
How should partners be positioning themselves and Cisco against the presumably strong AI story that a combined HPE-Juniper Networks will have?
It starts with the richness of the data that we have and how we’re leveraging it within each one of the portfolios, but across the portfolios as well, and what do I mean by that? That richness of data, if you look at it in say, security, we’re making that threat detection and resolution much richer and faster. But then when you [look at] security plus networking, for example, and you look at capabilities that we’re delivering as a result of that, things like firewall policies or firewall assistance, there are very few peers out there who have the assets that we have across the board, and then actually are able to bring the data at scale that we have to be able to deliver the insight. Maybe they can do it at a much smaller scale for very specific use cases, but I think what sets us apart from all of them is the breadth of the data that we have, the portfolio that we have covering networking, security, observability and collaboration, and now with Splunk coming in, how we can bring this [together] and also create a better experience for our customers as well. Our customers will tell you, regardless of whether it’s in security and observability, they’re not going down to one tool, but they want fewer. So, integrated capabilities provide more richness, but a better user experience, I think that is something where neither a Juniper [Networks] or an HPE can offer.
Cisco’s recent AI readiness survey found that only 14 percent of enterprises consider themselves ready for AI. How can partners help improve that stat?
I think channel partners, given what they see, can start by just helping our customers think through use cases. Because you need to start with: What is the use case that you’re going after? What are you using AI for? Just from an advisory standpoint, just [based on] what they see across multiple enterprises versus a single enterprise. The second thing is, they can absolutely look at [whether] this is something they want to create new advisory services around. Many companies struggle with, ‘I have lots of data and if I have lots of data, I can use AI.’ But all that data is sitting in different silos. So, there are advisory services that our channel partners can build [around] helping customers around their data strategy and execution of that data strategy as well. And it doesn’t mean that you build one giant data lake. It’s the use cases that you want to go after that helps you figure out what high-value data sets do you want to bring together, that you want to catalog, that you want to cleanse, that you want to normalize, and that gets you ready to be able to deliver those new services. That requires a tremendous amount of work across some organizations to be able to do that.
Why is it so important that partners grab on to AI and run with it in the market with Cisco solutions?
I would say the pace at which the technology is moving is just incredible. But on the other hand, you also want to be pragmatic about it. It’s great to be able to go out there and learn what’s new, but I think keeping pace with what your customers need right now [and] also giving them a view into what you’re developing that they could actually need in the future as well, I think Cisco does that best and we would like our partners to come along. So, it’s more of a, 'I want to be able to show the visionary capabilities and what we’re thinking, but I also want to meet customers where they are at that 14 percent.' Things like Motific [Cisco’s new SaaS platform for generative AI deployments within organizations] was absolutely meant for not just telling them, 'Hey, look at all these cool bits that we’re putting out there.’ We want them to be able to consume it. Things like, ‘Let’s talk about how we help you with defining the use cases.’ By the way, partners can say, ‘Hey, let’s just start with low-risk use cases and go out and learn these.’ These are things Cisco and partners working together can help our customers move beyond that 14 percent and be excited about what’s coming up versus it just being cool technology that’s out there. And I do think that this is a space that as much as hyperscalers have dominated, this next phase is going to be really about enterprises and it’s going to be about custom models with their own data [and] their own application logic.